I've problem installing OpenStack AutoPilot on vmware workstation.
I download OpenStack Autopilot test drive from ubuntu website. It's alright configuring Maas and adding nodes and about 80% of adding region. but in the task of adding units, it takes long time to infinity.
I do it for several time and that not pass this level and jobs remain In progress state.
I checked the log file for juju and I saw this error it could not connect to websecket in client with juju bootstrap.
for example my machine name is client1 and the job title "Add unit to 'mysql' service on 'client1.myautopilot'"
I check log file /var/log/juju/unit-mysql-0.log file, last lines of this file are as follow:
2015-12-19 10:58:58 INFO juju.worker.dependency engine.go:293 starting "api-caller" manifold worker in 3s...
2015-12-19 10:59:01 DEBUG juju.worker.dependency engine.go:301 starting "api-caller" manifold worker
2015-12-19 10:59:01 DEBUG juju.worker.dependency engine.go:268 "api-caller" manifold requested "api-info-gate" resource
2015-12-19 10:59:01 DEBUG juju.worker.dependency engine.go:268 "api-caller" manifold requested "agent" resource
2015-12-19 10:59:01 INFO juju.api apiclient.go:262 dialing "wss://client1.myautopilot:17070/environment/8a4d79c6-fa15-408c-817b-f0204b5a7783/api"
2015-12-19 10:59:21 INFO juju.api apiclient.go:270 error dialing "wss://client1.myautopilot:17070/environment/8a4d79c6-fa15-408c-817b-f0204b5a7783/api": websocket.Dial wss://client1.myautopilot:17070/environment/8a4d79c6-fa15-408c-817b-f0204b5a7783/api: dial tcp: lookup client1.myautopilot: no such host
2015-12-19 10:59:21 WARNING juju.worker.dependency engine.go:304 failed to start "api-caller" manifold worker: cannot open api: unable to connect to "wss://client1.myautopilot:17070/environment/8a4d79c6-fa15-408c-817b-f0204b5a7783/api"
2015-12-19 10:59:21 INFO juju.worker.dependency engine.go:351 "api-caller" manifold worker stopped: cannot open api: unable to connect to "wss://client1.myautopilot:17070/environment/8a4d79c6-fa15-408c-817b-f0204b5a7783/api" Please help me to solve this problem.
Thanks in advanced
1 Answer
I suspect the containers are not able to get an IP. These images were made for vmware's ESX/vsphere environment, where the vswitch has to be changed from its default configuration and allow promiscuous mode. I don't know if Workstation has a similar setting.
In essence, workstation needs to be configured to allow this "nested virtualization".
If you hit the cancel button in your deployment, Landscape will collect logs and allow you to file a bug with them attached. Or, if it cannot reach launchpad.net, it will let you download the logs in a tarball. Inside that tarball there is a juju status file which will show us the status of the containers.
